Functional scores are used to clinically assess total hip arthroplasty (THA) and for comparative purposes in orthopaedic literature. Previous pilot work has highlighted patient comorbidity as a potential source of bias in addition to the often-examined factors of age, gender and underlying diagnosis.

We analysed prospectively collected data relating to 217 primary THAs (Tayside Arthroplasty Audit Group database). Sample size was calculated from previous pilot data (significance level 5%, power 80%, effect size 0.25). Proportional stratified sampling was performed including all patient age groups (≤40, 41–55, 56–65, >65) and four fixation classes (cemented, hybrid, uncemented and Birmingham resurfacing). Five year Harris Hip Function Score (HHFS) was the dependent variable; age, sex, underlying diagnosis, BMI, pre-operative HHFS and comorbidity (Functional Comorbidity Index; range 0–8) were co-variates/co-factors.

In univariate analysis, FCI accounted for 11% of the variation in HHFS at 5 years. Regarding patients with FCI 0–1 as a reference group, patients with greater comorbidity had lower HHFS at 5 years (FCI 2–3: −3.95; FCI 4–5: −7.21, FCI ≥6: −6.92). In a multivariable model of HHFS at 5 years, FCI group, diagnosis, pre-operative HHFS, patient age and BMI were significant. HHFS at 5 years was significantly higher in FCI group 0–1 than any other category (2–3 P=0.006, 4–5 P<0.001, ≥6 P=0.002). In total, the model accounted for 29% of variability in HHFS at 5 years.

This is the first statistically robust study to examine the effect of comorbidity on THA function. These results strongly suggest that increasing patient comorbidity is associated with poorer THA function when measured using HHS.

Two significant conclusions can be drawn:

Patient comorbidity should be taken into account pre-operatively when considering potential future THA function.